Feeding behavior

Feeding behavior in fishes is both conservative and flexible. If it weren't flexible, we wouldn't be able to keep these fishes at all. To take an example, Pantodon buchholzi requires live insects in its diet; it isn't very flexible. If we don't feed it some live fruit flies from time to time, we may be disappointed in the African Butterflyfish and label it a "difficult" fish to maintain.
 
Feeding behavior in fishes is conservative. Fish would rather eat more of just what they were eating last meal. If you let them, they will train you to feed them only live food. Part of the amazing success of modern teleost fishes (they represent half of all living vertebrates) has been their ability to divide up available food resources among themselves by specializing. You are probably aware of the highly specialized feeding niches among the mbuna cichlids of Lake Malawi, where some teeth and jaws are specialized for plucking invertebrates out of narrow crevices, while other predators have become specialized eaters of fish scales or eye-biters. But the Amazon is also full of gastronomic specialists, according to John Lundberg, writing in Natural History for Sept 2001: Hypopthalmus, for example, eats tiny zooplankton; other fish eat snails (Megalodoras uranoscopus),  the scales of other fish (Catoprion), the tails of other fish (Magosternarchus), chunks of flesh and fins (piranhas), fruits and seeds (tambaqui), wood (Panaque), or blood and gills (Candiru catfish).
 
But many fish are more flexible generalists; they don't specialize in any one food source. A good example of one of these omnivores is the Dwarf Gourami, Colisa lalia: at the Colisa page, see the surprising gut contents actually found in Dwarf Gouramis raised in outdoor ponds.
 
About "excess" feeding. Fish that can swallow another bite, usually will. They eat like there's no tomorrow, because they have no sense of the future. Excess quantities of food, however, can't be assimilated. So a bonanza windfall passes largely unused through the fishes' digestive system. Aside from phosphate-based pH buffers, the prime source of phosphates in aquarium systems is in flake feed.
 
Feast and famine is the natural rhythm of predators. All fully-grown fishes thrive on what German fishkeepers call a "starve-day" once a week or so. You may notice, however, after a "starve-day"— or even a "starve-weekend"—  the fishes are noticeably less active, conserving energy, though perfectly healthy. 
 
In their natural habitats, fish can survive seasonal variations in the availability of any food at all. "After one dry-season fishing trip on the Rio Madeira [in Brazil, biologist Michael] Goulding analyzed 167 fish and found that not one had significant amounts of food in its stomach," wrote Catherine Caulfield in In the Rainforest, 1985 (paperback 1991, p 244). Of course the dry season does not coincide with the breeding season for most Neotropical fish, and partly for this very reason. D. Albrey Arrington of the University of Alabama and his team tallied 38,000 specimens, representing 254 species in North America, Africa and the Neotropics, and noted their stomach contents. Though grazers eat all the time and generally had full stomachs, the predators tended to have empty stomachs — up to 16% of the total inspected — especially the fish-eaters. Intensive brood-carers like cichlids also tend to dine less frequently. (This was in the August 2002 issue of Ecology, reported later in Science News.)